The performance of a digital implementation of an Applebaum-Howells type adaptive processor is analyzed for both a limiter and nonlimiter configuration. The performance is evaluated in terms of steady-state residue power, using either a single-pole filter or a perfect integrator to smooth the output of the correlation mixer. The latter filter is the more commonly used for digital implementations. It is shown that when using the perfect integrator filter for both the limiter and linear digital implementations, the steady-state average weight vector equals the optimum weight vector. Thus, for this filter, the steady-state residue power is the minimum possible for either implementation. When using the single-pole filter, neither implementation achieves the minimum possible steady-state residue power. The relative performance of the two implementations depends upon the relative gain settings. When the gains are adjusted to give comparable servo stability for the design maximum jammer power, a reasonable criterion for digital implementations because of analog to digital saturation, the limiter configuration always has smaller steady-state residue power.
Performance of Digital Implementations of an Adaptive Processor
IEEE Transactions on Aerospace and Electronic Systems ; AES-19 , 2 ; 168-173
1983-03-01
1609647 byte
Article (Journal)
Electronic Resource
English
A Digital Open-Loop Adaptive Processor
IEEE | 1978
|Kolmogorov-Chaitin complexity of digital controller implementations
British Library Online Contents | 2006
|Practical Implementations of Adaptive Wall Wind Tunnel Technology
British Library Conference Proceedings | 2008
|